Skip to main content

tv   Charlie Rose  PBS  December 3, 2016 12:00am-1:01am PST

12:00 am
>> rose: we can welcome to the program. i'm mark halprin of "bloomberg politics." we begin with charlie's interview with cardinal peter turkson in rome. >> we produce a small booklet which we call the vocation to have the business leader, and with that, invited business leaders to consider what they do in terms of their vocation. now, vocation for us, when people want to become priests or noneor -- nuns or whatever, buts to recognize themselves in the transformative role of businesses available as co-creators, as partners with god, in making the resources of nature serve the good of humanity. >> we conclude with charlie's
12:01 am
discussion of artificial intelligence with john kelly, head of research at i.b.m. and conversation that first aired on 60 minutes. >> charlie, most people think of watson as the game machine on jeopardy that most infamous match 40 years ago. but watson is more than that. watson is the beginning of a whole era of computing. when you think of computer which has almost a century of technology, started with mechanical switches that counted things, and then we moved to programmable systems that we told the system what to do and it improved our productivity. watson is the first of the next generation of computing built for big data and extracting understanding from massive amounts of data to help humans make better decisions so it's an entirely different computer. >> cardinal peter turkson in rome and artificial intelligence when we continue.
12:02 am
>> rose: funding for "charlie rose" has been provided by the following: >> and by bloomberg, a provider of multimedia news and information services worldwide. captioning sponsored by rose communications from our studios in new york city, this is charlie rose. >> good evening. charlie is away tonight traveling on assignment. i'm mark halperin of "bloomberg politics." cardinal peter turkson is one of pope francis' closest advisors. he will be running the new vatican office charging issues of migrants, the poor and those in need. he's also in charge of the dangers of capitalism in an era of climate change and global poverty. he spoke with charlie in rome
12:03 am
friday morning. >> rose: i have the great pleasure of introducing to you. i said to him, cardinal turkson, what do i call you? he said, how about peter? i won't use peter. i will use cardinal turkson. he is at the forefront of the ideas we have been discussing this morning as we've got upstarted as we have laid out the imperative of this conference. he has some familiarity with the united states, having been to school there. he went back to ghana, became an archbishop, then cardinal first nominated by pope john paul ii, then pope benedict brought him to the vatican, and pope francis has given him this new responsibility. here is what pope francis said to him when he handed over this responsibility which begins in
12:04 am
january 2017. "he will be competent particularly with regard to migrants, those in need, sick, excluded, marginalized, imprisoned, unemployed as well as victims of armed conflict, natural disasters and all forms of slavery and torture." you, sir, have your hands full. >> thank you. that's quite a case. first, good morning to all of you, very glad to be in your midst this morning to share these thoughts of pope francis with you. now, if governments were to deal with this h i don't know how many ministries they would have to deal with these issues, and it's all because pope francis, as part of his reform, he decided to bring the different offices. there used to be an office out there with migrants. there used to be an office with humanitarian assistance and then an office that dealt with
12:05 am
healthcare and sanitation. but i suppose to bring them all for a very good arrowhead for penetration and effectiveness, he decided to bring them all together under one head. so that's what we're prepared for. right from the beginning, we decided it should not be a conglomerate of offices. we decided to formulate a new vision of pope francis for the church's involvement in the social arenaa. when we're done we'll see what offices and structures we need to carry out the vision of pope francis. so that's what we're doing now. >> rose: you had a lot to do with the pope's insippicle on the environment. some consider you the primary architect connecting what happened with the environment and poverty. but i want to pick up again on this theme. tell me what you think of business. what do you think of business and what you think might come
12:06 am
out of this conference in terms of what business can do and what the church can do in a very concrete way in attacking questions of poverty and inequality. >> coconcretely, now, i came to the vatican in 2010. this was the midst of the financial crisis. and, so, we thought that it was our responsibility to provide a way of looking at this financial situation, and part of it was business, called banking and all of that. one of the first things we did was help and encourage the church to stop pointing, accusing fingers at the world of business. so we produce a small booklet which we call "the vocation of the business leader," and with that invited business leaders to consider what they do in terms of their vocation. now, vocation for us is for when
12:07 am
people wan want to become priesr nuns. it's transformative in their role as co-creators, partners with god in making the resources of nature serve the good of humanity. so we say god created a tree. god did not create furniture. it requires business to traction form trees into furniture, to transform minerals into what we use. essentially, that's what we think a business is. it's a partner with god in bringing the resources of nature to the concrete use of humanity and, therefore, business serves not only those who enable business to invest, but those
12:08 am
who need to benefit from business. so this is what we encourage business to do. sometimes we have to encourage businesses to recognize basically three things -- to ensure what itself is dignified, to insure wealth which they make is also good wealth, and to ensure that the customer, the relationship is also good. and we think when this is solved, then basically business is very dignified in our role in society. >> rose: how does business balance its responsibility to its stockholders with its responsibility to stakeholders, being humanity? >> stakeholders enable business to do what they do. they invest, bring in money to be able to do business. but the character of business
12:09 am
depends on more than those who bring i in the money. there are other people who live on the terrain. there are people whose lives will be transformed by the mining activity and who may be moved to one side or the other. so involved in this is more than, you know, investors. so we encourage more if you want a holistic view, you know, of business and its activity. at the end of the day, bottom line for us is that everything that happens should serve the well being of the human person. and what we say, therefore, is that the human person is the only thing god created for its own sake. everything else was created for the well being and to serve the human person. the human person was the one who was created not to serve anything else but for its own sake as it were, and when
12:10 am
business or any other human activity or engagement tends to make man serve another goal or aim, we want to remind everything should serve the well being of the human. >> rose: does the church believe that there is a necessary regulation and -- for example, it is not so much the market economy but the ideology that too often lies bind it, the dee fide market which resists the political oversight and regulation. what did the holy father mean? >> when you meet him tomorrow, i think he'll probably tell you what he meant. ( laughter ) but there is something from this we can pick up.
12:11 am
what i referred to when i got here, we put it in a small booklet and called it "reforming the financial system" and then went on to say, in the light of the global, you know, authority, global financial authority. the booklet was well received in several places and brought us at one point to frankfort to a bank to discuss with business people, investors, bank leaders and all of that. everything, the analysis we made of the crisis was accepted. we identified moral causes. we identified technological causes. but when it came to establishing global authority to exercise oversight of finance, there we had the greatest with we assistance. so the establishment of any form of authority to, you know, regulate this is not easy, and
12:12 am
probably that's what the pope is referring to. the second point, the only way we can ensure and guarantee that -- you know it so well, you know. every now it comes up the rates. a certain amount is necessary to ensure the ethical part of this. >> rose: in america they call it dodd-frank. >> okay. >> rose: the dignity of work. you've asked the world to consider the dignity of work and whether we ought to be about having machines perform the work of human beings. >> this, i think, is going to be a problem that would engage our
12:13 am
creativity and attention for quite a bit of time. since the days of pope john paul ii, we've been invited not to reduce work, but to recognize what it does to the human person. so we have been invited to recognize the objective and subjective character of work, not reduce to what we produce, but what also the worker and the human person. it ensure its dignity not because of the salary and it can put bread on the table of the family, but also creates an opportunity to exercise its own creativity, put to work his own talents and his god-endowed riches. so work, therefore -- the sense
12:14 am
of work is not to be what we produce, but to recognize that work also improves the subjective nature, character of the one who exercises work. the dignity then of the human person is made manifest also in what he produces or what he does. that's probably also the only way that any human person created in the image and likeness of god resembles god and produces things out of his own creativity, talents and endowments. and that's -- >> rose: for his own sense of dignity and the dignity of his family. >> yeah. >> rose: michael gerson, a columnist for "the washington post," said talking about globalization and the rise of populism and all of that, said, "the davo said in globalized elite, are leading participants in the economic system with the global chain, freely moving capital and rapid innovation that during the past 20 years has taken a billion people out
12:15 am
of extreme poverty around the world. this is arguably the greatest humanitarian achievement in history." without debating "the greatest, that's a remarkable achievement. >> certainly. >> rose: and i come back to it, what's necessary to make sure that government, n.g.o.s, business, with all the resources, with all the opportunities, with all the human capital that businesses have, the best and brightest, in many cases, how do you employ it along lines of morality and profit? >> the statement you just quoted of business and having lifted a lot of people out of poverty goes on to say, if we quoted from the same source, goes on to say, wh while it lifts so many t
12:16 am
of poverty, it also increases inequality. so it lifts out of quality but also inequality. and for us it would be great, if lifting people out of poverty doesn't increase inequality in any way. so what does it require? i think it is to be commended that business, you know, and trade and commerce and all of that help spread wealth in a very different way. an opportunity, occasion for people to exercise their own creative talents, to exercise work and all, but it would be great if, in the process, we are able to also make the spread of business and the spread of, you know, avenues of work and all that, help us also reduce the inequalities that ensue from this, so that, you know, i don't
12:17 am
know how many problem not to go into countries and i cite examples, in certain cases it introduces a lot of wealth but also widens the gap between the rich and the poor. >> rose: so the question is, could it have reduced -- could it have lifted so many people out of poverty without increasing the gap of inequality? could it have done that and what is the pathway to do that? and what can the church contribute? >> i think the part, as you observe, is not to make profit the main objective of business investment and all activities, but to make or establish tas main goal of business the ultimate lifting of people out of poverty and the well being of the human person. i know this requires investment,
12:18 am
but profit-making and all of that may not be the objective. as for what the church can do, i think the church can serve as a very decent outlet for promoting and lifting people out of poverty by way of direction, resources and investment to the needy areas where this needs to be exercised. and we have a case before us now. we now target looking at haiti, central republic south sudan, these are places which have poverty and we need to adopt a model that enables them to definitely and concretely emerge out of poverty. so next year one of our purposes would be not to see the path which has failed. i think we have the means to change how people live and do it
12:19 am
through housing, work and access to work. housing -- no, access to work is the same. so housing, work and, if you want, what we call land, okay, access to property, access to land they can use. when we ensure each one having a roof over their head and work they can do to sustain their lives and their family and ultimately something that they can call their own per time. >> mm-hmm. we wish we can do well. you know, i think there is a lot
12:20 am
of good will out there. >> rose: yes. there is a lot of show and display of good will to respond to the call of action, if you want, by pope francis. we've received guests from a fast food chain who have come to our office and said what can we do? we say, fine, begin by looking at your supply chain and try to improve in your supply chain. if you're able to do that, you would already have very many people. they come and say we recognize the access to energy, it's a problem in some areas, a lot of developing countries. so they want help with the production of a very big scale of solar system so that people don't struggle with smoke and lung cancer and stuff like that. that's something that we
12:21 am
welcome, and if it may be enough information, on account of this, the vatican accepted to participate in the exploring 2017. the team is new forms of energy for the future, and the vatican accepted to participate. suddenly, we go in there not because we can display new technologies about energy, but we're going there to try to tell a story about energy that is about the world being of humanity. so that's why we go there. about human life, and emphasize the positive side of lifting people out of poverty and conclude with something that normally is not considered, that's energy within all of us, and we identify that as spiritual energy that leads us
12:22 am
to pray, to meditate to do all the good things. >> rose: cardinal turkson, thank you for coming. >> thanks to you all. ( applause ) >> rose: here's what i want to know -- who is watson? what is watson? >> well, you know, charlie, most people think of watson as that game machine on jeopardy, the infamous match become five years ago. but watson is much, much more of that. watson is the beginning of a whole new era of computing. when you step back and think about computing which has almost a century of technology, started with mechanical switches that counted things, and then we moved to programmable systems that we told the system what to do and it improved our product iflt. watts b is the first of the -- watson is the first of the next generation of computing to
12:23 am
understand massive amounts of data to help humans make better decisions. it's an entirely different computer. >> rose: tell me about watson's intelligence. >> watson gains its swell generals from the data it's gathered or given. so it has no inherent intelligence as it starts, it's essentially a child. but as it's given data and outcomes, it learns, which is dramatically different from all computing systems in the past which really learned nothing. the more it interacts with people, the more data it gets, the smarter it gets. >> rose: what's its potential? is. >> its potential is unlimited because data keeps growing every day, it doubles every year. so the more data is fed, the learning continues to work and as it interacts with humans it gets smarter and never forgets. >> rose: this is almost like
12:24 am
you're watching something grow up. you've seen the birth, seen it pass the test, seen it get shatterer, you've seen it assimilate more. you're watching adolescence. >> that's a great analogy. on the jeopardy game five years ago, we put the computer on television, we let go of it and had no control of it. i felt as if i was putting my child on a school bus and i would no longer have control over it because it was going to start to learn. >> rose: it was reacting. it had no idea what questions it would get. it was totally self-contained. i couldn't touch it any longer and it's learned ever since. so fast forward from the game show five years ago. >> rose: how did you do it? we worked with the best people in the field, with the best computer scientists, but we work with the best oncologists
12:25 am
at sloa m.d. anderson and other. we ingest all the medical journals and watson learned over that time. >> rose: at its core, watson helps find information. >> no, it helps us make better decisions. the analogy i would use, charlie, is that, really, the search that we're used to today on the internet helps us find information. it will take a key word and find something related to it. >> rose: watson does what? watson goes through that information, but it understands the information. it reasons on that information. >> rose: reasons? easons on that information. it builds a model for what's being said. it understands relationships. it's not just looking for key words, and it will come back with an answer to a question or an observation that we as humans can't see. >> rose: but it is only as good as the information you can put in it. >> that's correct, and it can only learn if we give it more information. we have already consumed all the
12:26 am
world's medical literature. every day it gets updates on every clinical trial, every drug discovery, so it's in real time. >> rose: every medical document, every clinical trial, everything of importance to making a decision in the medical arena, the data is already there and watson has it. >> and it's updated every day. >> rose: so how does watson analyze it? >> so watson takes it in, and then it has a series of computer learning engines that tries to make sense of what it's seeing, and it's already built a model in its head, if you will, it's built a model of its world and as it gets new information, it assimilates that knowledge and tries to put it in categories much is same which as with we do. >> rose: how much control do you give it? >> we control through the data that we see. we do not give it infinite data in areas we don't want it to have, so we feed it data that's relevant that it will learn on. >> rose: people say when they
12:27 am
hear about artificial intelligence and this thing you've created called watson named after one of the i.b.m. founders -- >> yes. >> rose: -- and they say, how smart can it be, and can it be as smart as human beings? >> it:-- depends on your definition of "smart." it can be smart in finding information, reasoning on information and getting insights insights. bytes of information are too large. when i talk to doctors and lawyers, they always tell me, i'm in cognitive overload, i can't keep up with the information. i need a system to help me pore through that and help me reason. >> rose: so watson becomes their best friend. >> memorial sloan kettering doctors referred to it as their learned colleague. >> rose: how do you give watson the image?
12:28 am
>> it has a machine generated. >> rose: how did you make the decision as to what the voice would sound like? >> that was a very interesting choice. we could have given it any choice. we did a lot of market research ich was fairly neutral and hadce some degree of machine -- >> rose: you wanted it to sound like a machine? >> i had to sound something like a machine. >> rose: why was that important? >> well, we wanted to represent what technology could now do. man being or human knowledge.e a >> rose: put hair on top of it and make it look like a human being. >> that's right, it's still a machine. >> rose: does it have personality? >> today, it has no personality but, interestingly, it can understand your personality because it can take your language and it can understand the words and the patterns of your speech and, based on that and its knowledge of psychology, it can build a model. >> rose: so it can do a behavioral profile of me, if i
12:29 am
input into it things that i do. >> that's right, based on the words you use and patterns of your speech, it will develop a psychological profile. >> rose: is that good in medical stuff? >> it is. as we get into neurological diseases and diseases of the brain, we are finding watson can pick out pat, of speech which often are early indicators of neurological problems. >> rose: that's crucial, isn't it? >> crucial. >> rose: early detection. that's right, opens up a whole new set of possible therapies and delaying or curing disease. >> rose: how does it do that? it listens very carefully to the words you use and the patterns and it turns out that every disease will start to affect your speech. it's often said that the eyes are the windows of your mind. actually, your speech is a tremendous window into your mind, and often neurological diseases manifest themselves into speech patterns first. >> rose: in fact, people who have a.l.s. have said to me, they first noticed it in voice. >> that's right, and we've
12:30 am
looked at areas like schizophrenia and we can clearly see difference in speech patterns between people with schizophrenia and normal speech. >> rose: so personalities, but then there is ethics and morals. >> it only knows what we teach it. >> rose: so you can teach it ethics? >> it only learns our ethics. it doesn't develop our own ethics. it can look at our patterns of information, our conclusions and begin to mimic that but it can't be developed. >> rose: is there any reason for it to be an ethical or moral machine? >> no reason for it to be one or the other. it's still a machine. >> rose: what doesn't it have that you wish it had? >> well, you know when we did the original jeopardy match, it did one thing well, it understood natural language and could understand open-ended questions, but at that time had no ability to understand images. if you put an image in front of it or feed it a digital image,
12:31 am
it had no idea what it was. since that time, we've taught watson basically how to see, how to analyze images. that's so important because so much of the world's information is in images. three-quarters of the information in data in healthcare now are in images. >> rose: because we see with our brain. >> it's about the brain, not the eye. what watson does, it looks at images, we tell it it's this or that, and it learns on its own to go forward and understand images. we tell it what it is to begin with and it will extract it's own learning. >> rose: is that the way you teach it, by the information you give it? >> yes. that's what's so different about this new era of computing. in the past we would have to go in with programmers and tell the machine to do something specific
12:32 am
differently. in this case you don't reprogram watson. you give it new information. it develops and reasons in new ways and gets new insights on its own and working with humans. >> rose: when did you make the decision that was the way to go? >> in the early 2000s, we were looking at, at the time, what was artificial intelligence, and everyone before this effort had tried to build systems that directly mimic human understanding. >> rose: human learning. human learning. and they were developing all sorts of rules. they were trying to program systems to be like their brain, and there were some tremendous breakthroughs in ibb mmm research that said we're not doing it that way, we're doing the it entirely differently, purely statisticsle. we're not going to tell the system, we're going to put images in and give ability to
12:33 am
learn. that was a tremendous breakthrough in technology and will be in the history of computer. >> rose: is that the decision you based the company on? >> that's it? you basically said to i.b.m., this is our future right here? >> that's it. we fundamentally made the decision this is about helping humans make better decisions. >> rather than trying to imitate and -- >> and displace or whatever. >> rose: why is that the better way to go? >> we've proven time and time again that man plus machine will beat either machine or man. >> rose: so this is not man versus machine for you. >> no. >> rose: it's man plus machine. >> exactly. >> rose: any doubts about that decision? have you been confirmed in that decision by everything that happened since you made it? >> yes, and we see cases in every industry, in every domain, now, charlie, where that combination of the watson system plus an expert or someone learning in that field can make
12:34 am
a better decision in a more timely manner. think about a medical doctor who can't deal with all that information and yet only has a few minutes to prepare for a patient. bring that capability together, now the doctor can make a better decision with the patient. >> rose: what surprised you in this? >> there were many times in the early days when we were bringing up the watson system that it would answer a question, charlie, and i was stunned, and it was how did it do that. and we would look at all the computer traces to figure out how it did it. i was shocked -- i thought i was shocked and surprised at that time, but when like at what watson has done in areas like healthcare in the last three years, it's basically gone to medical school, through its residency and has become an expert in some forms of cancer in three, two and a half years. >> rose: your child has gone to medical school and now is
12:35 am
through. >> it's through. >> rose: do you feel a certain paternity about all of this? >> i'm proud of the team at i.b.m. and the i.b.m. company who made the investments to create this. now it's bigger than i.b.m., artificial intelligence, cognitive machines is a whole new field that's sort of resurrected in the i.t. industry and i think it will transform i.t. and healthcare and all the other questions in society. >> rose: the other thing you have been witness to is the cultural shift. in terms of artificial intelligence, and everybody was instantly attracted to the idea, and then there was a lull. >> that's right. >> rose: and now every company that i know that's a tech company in the forefront of their business is thinking about mponent of their business. >> that's right. that's right. >> rose: you watched this happen in a short time! >> in a very short time. you know, people have studied artificial intelligence in the
12:36 am
1940s and 50s and failed. the best companies were trying to approaches and failed. the breakthrough here was the try an entirely different approach. i think the jeopardy watson stunned people, and now that they see the progress it's making, it's attracting tremendous attention. >> rose: take me to the jeopardy challenge and what it was and what it meant and how you handled it. >> i think it was a tremendous inflection point. >> rose: because you had already done chess. >> we had done chess. >> rose: and you learned something from that. >> that's right. you know, these games are often viewed as man versus machine, but we're trying to focus our technology on something that's a milestone. so the goal wasn't the jeopardy match. that was just a milestone and a demonstration of its capability. >> rose: i want to know about chess, first, what you learned about that. >> sure. >> rose: because i was shocked
12:37 am
you could beat carey. >> before that time chess was viewed as too many options, and we built a very large computer that basically did very deep searches of all possible moves faster than gary kasparov could do it. and we went on to beat him. but that was to tackle a deep search problem. >> rose: right. watson is aimed at an entirely different problem which is the world of data, now, is so large, and people cannot deal with it, they cannot extract knowledge quickly enough, so we're trying to improve decision-making versus do deep searches in a game. >> rose: so the jeopardy challenge represented what? >> we felt the jeopardy challenge was interesting because it tackled the natural language problem, right. natural language is very
12:38 am
difficult for a computer system. >> rose: what do you mean by natural language? >> our speaking language. our written english. very difficult for a computer. it likes to be programmed in computer languages, which is what programming is all about, but we wanted to teach this system to understand first english and other natural languages because so much of our knowledge, charlie, is in our natural language written in books. now it's all digitized. so we felt that the first milestone was to really crack the natural language and to be able to answer any question in any field quickly better than those two human beings, and that was quite a challenge. so that was a major milestone. but we were looking beyond just human language into images, signal processing, other sorts of data. >> rose: how long did it take you to get ready? >> you know, we started about five years before the match, and at that time, it was taking watson hours to answer one
12:39 am
question and it never got it right. but i knew about 12 to 18 months before that match that we were going to be head to head with those human beings, and it was going to be a very close match at that time. i remember, when i introduced the match to the audience that day, i said to the audience, you know, this is not a question of if, this is only a question of when a machine will be able to win at this game. i don't know if we'll see it today, but we're going to see it very soon, and, of course, we went on to see it? there was a story i read about a bunch of people at an event and there was a television on in the other room and they all rush in there because they want to watch jeopardy. >> yes. >> rose: that taught you what? so we were looking for -- we love grand challenges in i.b.m. what's a problem that a system has never done before?
12:40 am
>> rose: you're here to solve problems. >> we had a bunch of researchers in a restaurant and it was a time when jeopardy was on a tv at the bar and my assistant got up, ran in there and said if we can build a machine to do that, we can build a machine to transform industries, and that became the milestone and the next goal for the team. >> rose: that's a big day, isn't it? >> yeah. >> rose: the big day that you realized, if we can do this, the future is hours. >> that's right. the day we realized, if we can do that, the future is ours, and hen the day that the team said we're going to do it this way versus old traditional ways. >> rose: was it a sure bet? not at all. not at all. not at all. >> rose: you know how to access all the world's information. >> everyone had failed at this before. eryone had failed. for decades. >> rose: why had they failed? because they had tried to basically, in structured ways, re-create the human mind and
12:41 am
re-create human intelligence, which we still have no idea how the to do. >> rose: they felt because they didn't appreciate man plus machine. >> that's right. and our goal was only to help humans make better decisions. >> rose: so tell me about the final jeopardy challenge. i mean, your heart was beating faster. maybe you did all this work for naught. >> right. charlie, the day before, we could have told watson to don't embarrass i.b.m. or try to play with the other two players or play to win, no matter what, and we told watson play to win. >> rose: how do you tell watson to play to win? >> you tell it to basically bet more money, get more aggressive in the game, no matter what, to try to win. so you can imagine what the outcomes could have been, had we fallen behind. >> rose: but i'm fascinated by you told. it's literally like you're
12:42 am
sitting there, it's a boxing march, and you're telling your fighter, play to win, go for the knockout, win this thing. >> that's right. it was big for i.b.m. because it was our centennial year in 2011, but more importantly, it painted the direction for the i.b.m. company and more importantly for our industry, where are we going to take tensionology in the future. it's not just about gadgets or cell phones. it's about how will we make humans make better decisions, transform industries, solve the healthcare and climate problems, et cetera? while winning the game felt good, it was a confirmation we could change the world with this technology. >> rose: you really believe you can change the world? >> change the world. >> rose: change every industry? >> that's right. in fundamental ways, we are going not only to take the cost out of the healthcare system that's almost bankrupting countries, we're going to prove the outcomes. these systems are going to help doctors save lives. >> rose: you know the people
12:43 am
that doubt you. >>yes. >> rose: why do you think they doubt you? >> i think they doubt because this technologies like this have failed in the past, but i think they -- while they doubt, they watch us very carefully. recently there is been an explosion in the tech industry over artificial intelligence. >> rose: everybody wants to. everybody wants to. >> rose: how do we gear up? ight. >> rose: if this train is leaving, i want on it. >> this is one, charlie, you cannot miss. the other point i would make is because its a learning system, the longer you way to get on the train, the further behind you will be, because that machine is learning every day and getting smarter and smarter and smarter. >> rose: smarter today than yesterday and not as smart as tomorrow. >> that's right. >> rose: you're saying you have to get on board now. >> yeah. ere you let the early adopters
12:44 am
do it, wait till it's debugged and get on the train, because the train will be way out of the station and is accelerating because we're feeding it more data. so this is one where you want to be an early adopter, get on and you want the machine to learn in your domain, the earlier the better. >> rose: in the evolution from jeopardy to fighting cancer, what was necessary to go that distance. >> three to five years to get it done. the first thing is watson had to understand the language of healthcare and the language of cancer because it is different. >> rose: one thing watson had to do was understand natural language. >> that's right. >> rose: and ten has to understand the language of cancer. >> of cancer, which is very sophisticated. it then had to ingest all the information from the medical literature about cancer, then we had to have top doctors at
12:45 am
memorial sloan kettering and m.d. anderson to prime the pump of learning and get watson learning on cancer. >> rose: m.d. anderson, sloan kettering, a lot of people identify sort of the cutting edge in understanding cancer, institutions like those two and others. >> that's right. >> rose: when you came to them with the possibility of this, what was their response? >> interestingly, they came to us. they came to us. the top docs at memorial sloan kettering said i've never seen a tool like this, i think we can do this with watson. i think we can go after cancer with watson. i think we can scale our knowledge, the best in the world, not just to make the very best at memorial sloan curing or m.d. addressem.d. anderson betto help those doctors who don't have access to leading edge
12:46 am
technology and education. this system has been trained by the best, learns all new relevant information every day in seconds, and doctors have access to the knowledge. >> rose: can it make a misdiagnosis? >> watson is never perfect. it has the ability to believe it's more right than wrong, but it's never perfect. i mean, it can be very, very precise and very, very accurate, often better than human beings. >> rose: often? well, in cases where watson has had sufficient data and sufficient time to learn, it will be as good or better than the best because it has more data that it can ingest than any human being. >> rose: success is measured by return on investment? >> yes, it always is in business. but i think beyond that, this may be what separates i.b.m. from many is that we want to innovate, obviously, from a
12:47 am
business that's core strategy. we want to help society. we chose healthcare as the first place to sort of aim watson. yes, it's a big industry and being digitized and yes we want to have a big business there, but we felt that the impact that we could have on human lives was better than any other industry so we staed there. >> rose: you have to be careful because you can't overpromise. >> that is correct. but the potential, charlie, of this. you know, i have been in this industry now for over three and a half decades, and i've built some of the world's largest super computers. i was involved in the system that built gary kasperov. this breakthrough, this technology is like nothing i've seen before in three and a half decades in this industry and what's most exciting to me is
12:48 am
not that it's a bigger, faster computer, it's going to transform industriesreich healthcare. >> rose: you know people at google think deep mind will do this. >> they're working with aim a.i., they're applying it to their own business model. they've had great successes. they took on the game of "go" recently and won. i congratulate them. >> rose: that's more complex than jeopardy. >> no, i would say it's more like the chess game. it's a series of moves, more complicated, in the sense there is more potential moves than a chess game, but you're basically looking at combinations of steps. that's a different problem and, again, i congratulate them. >> rose: could you have beaten "go"? >> we didn't take that probable on. >> rose: but could you? if we had focused on that problem, we could have. but that was not our goal. >> rose: one game at a time. we wanted to go after the big data problem. we wanted to go after industries. we wanted to help humans that were basically in this cognitive
12:49 am
overload because of information, and we wanted to help them make bet decisions. >> rose: is any of this scary to you? >> not at all. >> rose: all-owferl artificial intelligence, knows where we are, what we're doing, what we're thinking and some suggest one day may control us. >> it only knows what we want it to know, what we feed it. >> rose: and is it possible that we can make it so powerful that it's more powerful than the person? >> there are certain attributes that what we humans are that watson will never be that. >> rose: like what? every technology, people have fear of what it can do -- electricity, locomotive trains, oh, my god, they will accelerate forever, we'll be in trouble. the way i look at it is what is the cost of not pursuing this? there is no other technology
12:50 am
that's going to be able to possibly help us find better cures for cancer, treat patients better than a system like watson. we must pursue this technology, and it would always be limited by the information we feed it and the interactions with humans. >> rose: why do you think people like elon musk, bill gates, just to name two, and i know a lot of other people have reservations, perhaps not to the extent they do. >> some people still think as the old idea of artificial intelligence as trying to re-create the human. >> rose: people keep up with it. >> they keep up with technology, but if you really understand a system like what watson learns and what it can't learn without certain information, you get very comfortable. >> rose: what tech changes will make watson better than it is today in the next five years? >> the underlying technology that we've now brought from this one box machine that competed on jeopardy up to the cloud has been a tremendous advantage
12:51 am
because, now, we can go global with watson. we can access watson any place on the planet instantaneously through the network and the cloud. that's been a tremendous breakthrough. but still the quantity of data that's been generated by human beings is surpassing even what watson's underlying technology can keep up with. >> rose: so? so we need real breakthroughs in underlying microtechnology and break thrust in how to structure this much data. we need new mathematical algorithms, more new technologies that can find more and more subtle information. >> rose: so watson has put a demand on everything that is known now and almost saying to go further you have to get better. >> we have to get better at every level of the system to advance watson or even watson won't be able to keep up with the amount of data we as humans generate and the machines we're creating.
12:52 am
>> rose: in that arena what excites you the most, the arena of what watson is demanding you to do in order for it to do what it's capable of doing, given updating, modifying, improving, creating new? >> the merging of data from different sources is producing whole new insights. one example, we're doing work with metronic, which is a medical device company, fascinating, around the disease offdiabetes, and we've taken what watson's learned through the medical literature, the zillions of lives we have access to, information on diabetes, and we bring all the information together and for the first time ever, watson was able to predict with high precision the onset of
12:53 am
a diabetic hypoglycemic attack hours ahead of time. the state of art is an alarm goes off when you're almost in critical condition. >> rose: almost but not. but not, but you better get to the emergency room instantaneously or you're going to have severe problems. we now can predict that hours ahead of time. >> rose: right now the focus is on healthcare. where else would you be going? >> we're starting to look at literally every industry. obvious ones are financial services where we begin to look at financial transactions and things like that. >> rose: you can match the velocity -- >> because humans can't keep up with trades and things now. another one that's very interesting that we're starting to work in is in education because we as humans, just as we all have different medical conditions, we all learn in different ways. we learn in a very personalized way. think about watson, now, looking
12:54 am
at how we as a child are learning or not learning and adapting the information to tune it for how we learn. i think it's going to accelerate learning, and we're very excited looking at pre-schoolers because we know so much of the learning occurs before you're five years old. think about watson now helping us learn or early detection of autism or other childhood diseases. >> rose: you look at the potential of watson today. is it at 10% of its potential? 25% of its potential? 50% of its potential? >> oh, it's only at a few% of its potential. >> rose: 1, 2, 3? yeah, i think this is a multi-decade journey we're on with this kind of technology, and we're only a few years into it and just starting to learn about the capabilities of these
12:55 am
systems. so i think we're in a few percent of the ultimate potential of this. and the way this learns and the way we're generating data, the learning, the capability of this is on an exponential curve. >> rose: it is a giant voyage into the new world. >> also, and we're at the very beginning. >> rose: for more about this program and earlier epi
12:56 am
captioning sponsored by rose communications captioned by media access group at wgbh access.wgbh.org >> rose: funding for "charlie rose" has been provided by: >> and by bloomberg, a provider of multimedia news and information services worldwide. >> you're watching
12:57 am
12:58 am
12:59 am
1:00 am
♪ this is "nightly business report" with tyler mathisen and sue herera. help wanted. companies are hiring, the unemployment rate is falling. but not everything is going gangbuster. strategic advice. donald trump picks a who's who of corporate america to advise him on the economy. but there's one group noticeably missing. girl power. female action figures are invading the toy aisle, thanks to the bright idea of a mom turned entrepreneur. those stories and more tonight on "nightly business report." for friday, december 2nd. good evening, everyone. and welcome. 74 straight months of job creation. employers continue to add workers at a steady clip last month, and the unemployment rate dropped to a nine-year low. according to the labor

43 Views

info Stream Only

Uploaded by TV Archive on